Going Public in Poland: Case-by-Case Privatizations,
Mass Privatization and Private Sector Initial Public
Offerings
By: Wolfgang Aussenegg
Working Paper Number 292
December 1999
Going Public in Poland: Case-by-Case Privatizations, Mass
Privatization and Private Sector Initial Public Offerings
Wolfgang Aussenegg*
Vienna University of Technology Department of Finance
Address: Favoritenstraße 9-11, A-1040 Vienna, Austria, Europe E-Mail: waussen@pop.tuwien.ac.at.
Phone: ++43 1 58801 - 33082 Fax: ++43 1 58801 – 33098
First Draft: July 1999 This Draft: December 1999
* I am grateful to Hans Reitbauer for helpful comments and suggestions. Financial support of the
Vienna University of Technology in acquiring data used in this study is gratefully acknowledged. The author also likes to thank Reuters Ges.m.b.H (Vienna) for providing data.
Abstract
This study compares the characteristics and the price behavior of case-by-case privati-zation initial public offerings, private sector initial public offerings and the mass priva-tization program in Poland over the first eight years after the reopening of the Warsaw Stock Exchange in April 1991. There is evidence that the Polish government is market-oriented in the sense of Perotti (1995), trying to build up reputation for its privatization policy over time by underpricing and selling a higher fraction at the initial offer. In the long-run private sector initial public offerings tend to underperform whereas case-by-case privatization initial public offerings experience neither an under- nor an overper-formance. What is surprising in this context is the significantly negative long-run after-market performance of the certificates and shares of the mass privatization program.
Keywords: Poland, Privatization, Initial public offerings, Underpricing, Long-run performance
Summary
In the past decade a lot of state enterprises have been privatized in nearly every Central and Eastern European transition economy. Poland provides the unique opportunity to compare the short and long run price behavior of three groups of firms listed on the Warsaw Stock Exchange: (i) Case-by-case privatizations via a public share offer, (ii) private sector initial public offerings and (iii) certificates and shares of the mass privati-zation program. The results for the sort run price behavior reveal that all three groups are significantly underpriced (investigation period: April 1991 - December 1998). The underpricing of case-by-case privatizations and private sector IPOs can best be ex-plained by investor’s demand: A higher demand for an issue leads to a higher under-pricing. In accordance with the political uncertainty hypothesis (Perotti, 1995) the un-derpricing of case-by-case privatizations is higher if a higher fraction of the capital is sold at the initial offer.
The long-run performance of Polish case-by-case privatizations is neither (significantly) positive nor negative over the first three years of aftermarket trading. This non negative long-run performance is consistent with a market-oriented government trying to build up reputation for the privatization program in the course of time. On the other hand, the long-run performance of private sector IPOs tends to be negative. This is in line with the evidence for many developed markets. The three year aftermarket performance of the certificates and shares of the mass privatization program is even worse than the negative performance of private sector IPOs. This is surprising as Polish government officials tried to design the mass privatization program in a way to create an efficient corporate governance structure. One reason might be that there are too few incentives for the management of the privatized firms to restructure efficiently.
Introduction
In the past two decades a lot of state enterprises have been privatized in the world. This applies to developed economies like the UK, France or Germany as well as a lot of de-veloping economies in Asia, South America or Central and Eastern Europe. As Perotti and van Oijen (1999) show, by reducing political risks privatization programs in emerging economies are able to provide an essential part for the development of the local capital market and the development of the economy of the corresponding country.
Two types of privatization methods can be distinguished: (i) Case-by-case privatization, in which state enterprises are sold case by case (one after the other) and (ii) mass priva-tization programs (MPPs), in which a multitude of firms are sold at the same time. In most cases the aim of MPPs is to transfer a big number of state enterprises into the ownership of the citizens. They are characterized by the fact that every citizen is entitled to receive a particular number of certificates (vouchers) free of charge or at a very low price.
This study examines the going public process in Poland. After a break of more than 50 years due to communism the share trading on the Warsaw Stock Exchange (WSE) started on April 16th, 1991 with five privatizations (Tonsil, Prochnik, Krosno, Kable and Exbud). April 16th, 1991 is also the start of the investigation period of this study. In contrast to other Central and Eastern European countries, like Russia, the Czech Repub-lic or Slovakia, Poland decided to start the privatization program with case-by-case pri-vatizations as a first step. Only four years later a mass privatization program (MPP) was started.
The aim of this paper is to investigate the price behavior of three groups of firms listed in the WSE and to analyze differences between these groups. The three groups are: (i) Case-by-case privatization initial public offerings (IPOs), (ii) private sector IPOs and (iii) certificates and shares of the MPP. For every firm and the MPP the analysis of the price behavior starts with the subscription period and the issue price and ends three years after the first trading day on the WSE. In a first step the initial return and the
un-derpricing are documented and possible reasons for differences in the unun-derpricing lev-els are examined. The analysis of the latter uses classical explanation modlev-els, which should be of importance for the private sector IPOs as well as models especially de-signed for privatizations (see Perotti (1995) and Biais and Perotti (1997)). In a second step the long-run performance for the first three years of aftermarket trading is exam-ined.
In contrast to conventional (private sector) initial public offerings (IPOs)1 empirical studies investigating the characteristics and the price behavior of public offer share pri-vatizations have emerged only recently. One of the first studies in this context was car-ried out by Perotti and Guney (1993). They documented that privatization programs have strong regularities across countries. Governments mostly sell only a part of a state enterprise at the initial offer and retain a significant stake for a long interval of time.
Menyah and Paudyal (1996) investigate the initial return and the long-run performance of privatization and private sector IPOs in the UK. Privatization IPOs offer a significant (market adjusted) underpricing of +38.70%, compared to 3.48% for private sector is-sues. The long-run performance is also better for privatization IPOs with a significant buy-and-hold abnormal return (BHAR) of +60.97%, in contrast to only +3.01% (not significant) for private sector IPOs. A comparison of privatization and private sector IPOs in Malaysia is provided by Paudyal et al. (1998). Privatization IPOs are signifi-cantly more underpriced than private sector IPOs. The market-adjusted long-run per-formance over the first three years shows no significantly positive or negative perform-ance for both privatization and private sector IPOs. Dewenter and Malatesta (1997) ex-amine the initial returns for privatization and private sector IPOs in eight countries. On average there are no significant differences in the underpricing of these two groups. While UK privatizations experience a significantly higher underpricing than their pri-vate counterparts, in Canada and Malaysia the underpricing of privatizations is signifi-cantly smaller.
1
Jones et al. (1999) examine a 59-country sample of 630 privatization IPOs (share issue privatizations) during the period 1977-1997. Their results show that privatization IPOs are significantly underpriced (mean: 34.1%, median: 12.4%). A comparison to private sector IPOs (see e.g. Loughran, Ritter and Rydqvist (1994)) shows that their docu-mented underpricing is approximately similar to conventional (private sector) IPOs. Jones et al. document that the underpricing in privatization IPOs is influenced by sev-eral factors. The underpricing increases if a high fraction of the share capital is sold at the initial offer. Therefore, if a government wants to sell a larger stake, the issue price is fixed at a lower level, which results in a higher underpricing. In addition they document that the underpricing level increases with a country’s income inequality and with the size of the firms.
The long-run aftermarket performance for a 36-country sample of 264 privatization IPOs (period: 1981-1997) is examined by Megginson et al. (1998). Regardless of which benchmarks are used, the performance for all holding periods (one, three and five years) is significantly positive. In addition, privatization IPOs making a seasoned equity of-fering (SEO) do not underperform (and not overperform) in the long-run. This is in contrast to e.g. the evidence for the US (see Loughran and Ritter (1995)), where compa-nies after a SEO significantly underperform in the long-run.
Several reasons are provided in the literature why the long-run performance of privati-zation IPOs might be better compared to private sector IPOs: (i) The informational asymmetry about the value of the firm is lower than for private sector IPOs. Privatiza-tion IPOs are often big companies and well-known by the populaPrivatiza-tion. (ii) A government wanting to carry out a privatization program over a longer period will do its best to en-sure that their privatization IPOs are going to be an economic as well as political suc-cess. In this context it is especially important that people get a positive view of the gov-ernment’s privatization program in order to attract domestic retail investors again for future privatizations. (iii) Private sector IPOs are often sold after an especially good operating performance, which results in a negative long-run performance.2 For
2
zations there exists multiple evidence that the financial and operating performance sig-nificantly improves after privatization (see Megginson et al. (1994), Boubakri and Cos-set (1998) or D’Souza and Megginson (1999)).
Several models have been designed to explain the price behavior of private sector IPOs. The most notable among them are the winner’s curse model (Rock (1986)), signaling models (see Allen and Faulhaber (1989), Grinblatt and Hwang (1989) and Welch (1989)), the ex-ante uncertainty hypothesis (see Beatty and Ritter (1986)) or the reputa-tion hypothesis (see e.g. Carter and Manaster (1990)).
The price behavior of privatizations have especially been modeled by Perotti (1995) and by Biais and Perotti (1997). Perotti (1995) distinguishes in his model between two types of governments: market-oriented and populist governments. A market-oriented govern-ment is characterized by the fact that it wants to privatize state owned enterprises seri-ously and irreversibly. This does not apply to populist governments. Whether a govern-ment is market-oriented or populist can only be realized by the market if there are con-crete signals. Only announcing to be a market-oriented government is not enough.
What signals can a government send out to be recognized as market-oriented? Perotti (1995) shows that if the uncertainty about the future privatization policy is high (this applies especially to emerging markets) a larger fraction has to be sold at the initial of-fer. Otherwise the market would get the impression that the government possibly does not want to give up control rights over the state enterprise. Depending on the level of political uncertainty the offer also has to be underpriced.
A market-oriented government is willing to accept lower gross proceeds due to the un-derpricing. The economic advantage of the privatization is more important for it than the disadvantage of lower gross proceeds. In contrast, a populist government is not willing to accept a (high) underpricing and the corresponding lower gross proceeds. A populist government would therefore also try to sell a not too small fraction, but at a maximum issue price (and therefore minimum underpricing). A market-oriented gov-ernment with difficulties in signaling its identity can, in addition to selling a larger
frac-tion at a lower price, equip the privatizafrac-tion IPO with a fixed issue price and distribute a larger portion of the offer to domestic retail investors and employees (see Biais and Perotti (1997)).
In addition Biais and Perotti (1997) document that a privatization program can be suc-cessful if median class voters receive a sufficient amount of shares. As a reversal of the privatization program decreases the value of the already privatized firms, median class voters will support the privatization efforts of the government, which results in a sup-port of the government in elections. In order to attract median class voters to buy enough shares so that their political preferences are similar to those of the government, underpricing is necessary in most cases. The more income inequalities there are in the population, the poorer are the median class voters and the higher the underpricing has to be to persuade them to buy shares.
This study contributes to the existing literature in several ways:
(1) An Eastern European emerging market is examined completely from the reopening of the stock market. Other studies, like Jones et al. (1999), Megginson et al. (1998), Dewenter et al. (1997) or Huang and Levich (1998) investigate several markets from all over the world as well as for developed and emerging markets jointly. Their re-sults are of high importance but they are often averages over all the countries under investigation. It is also of interest to examine testable hypotheses for a single (East-ern European) emerging market. The investigation of a single market can provide additional insights as the database is more homogeneous. This is valid especially re-garding listing requirements, the legal framework as well as the political and eco-nomic environment.
(2) A share privatization program is examined right from the beginning. This allows e.g. testing to what extent a build-up of reputation concerning the privatization pro-gram takes place over time and whether the hypotheses of Perotti (1995) and Biais and Perotti (1997) apply to Poland.
(3) In addition to privatization IPOs also private sector IPOs are examined and both groups are compared to each other. This allows investigating the differences
be-tween privatization and private sector IPOs. Until now there have only been a few comparisons of this kind in the literature. Examples are Menyah and Paudyal (1996), who examine privatization and private sector IPOs for the UK and Paudyal et al. (1998), who compare these groups for Malaysia. The studies of Huang and Levich (1998) and Dewenter and Malatesta (1997) also compare privatization and private sector IPOs but they only analyze the underpricing. This study is therefore the first one to compare the long-run performance of privatization and private sector IPOs for an Eastern European emerging market.
(4) Finally, the initial and aftermarket price behavior of the case-by-case privatization program and the private sector IPOs is compared to those of the mass privatization program (MPP). Such a comparison has not yet been made in the literature.
The following section describes the database used and provides descriptive statistics for the IPO groups. In the third section the empirical evidence regarding the initial returns is presented and possible reasons for the underpricing are examined. Section four deals with the aftermarket performance and section five gives a summary of the findings.
2. Data
In Poland shares are traded on the Warsaw Stock Exchange (WSE), which was founded in 1817. During the Second World War and the following decades of communism the WSE was closed. The WSE reopened on April 16th 1991 with the listing of five firms. From April 1991 till the end of 1998 altogether 187 firms went public with a listing on the WSE. 101 firms started trading on the main market of the WSE, 68 on the parallel market and 18 in the free market. In total only four firms were delisted (three mergers and one going private). At the end of 1998 therefore 183 firms were listed on the WSE: 117 in the main market, 51 in the parallel market and 15 in the free market. The differ-ences to the number of going publics can be explained by changes from one market segment to another. The parallel market started on 22.4.1993 and the free market on 20.2.1997.
This study uses the two biggest market segments: the main and parallel market. They comprise the biggest and most liquid firms trading on the WSE. The investigation pe-riod starts with the reopening of the WSE on April 16th 1991 and ends on December 31st 1998. Of the 169 firms going public in the main and parallel markets only those are used which (i) made a public offer before trading on the WSE and (ii) for which all nec-essary information (i.e. especially issue price, issue volume, demand multiple, contract type (fixed price or tender), rationing information, dividend payment, seasoned equity offering information) were available. Therefore the final sample comprised 149 firms (93 from the main market and 56 from the parallel market).
To build up the database the following sources were used:
(i) Historical Database of the Warsaw Stock Exchange (WSE): daily stock prices, divi-dend payments, number of outstanding shares, price-earnings ratios, first trading day on the WSE;
(ii) WSE Fact Book, several editions: issue prices, first trading day, ownership origin of listed companies, dividend payments, stock splits, mergers, delistings.
(iii) REUTERS Business Briefing Archives:3 Subscription period, issue price, owner-ship structure, demand multiple, rationing data, gross proceeds, seasoned equity offer-ings, stock splits and other company related information.
All information received in (i), (ii) and (iii) was checked and completed using informa-tion available of the Polish stock market in Reuters 3000 Equities History as well as in Datastream.
For investigation purposes the database of 149 firms was divided into two parts: (i) pri-vatization IPOs and (ii) private sector IPOs. In the sample of Privatization IPOs all those firms are included which are classified as Public Offering by State Treasury in the
3 Comprehensive business database, with access to national and international news wires, news
Fact Book 1999 of the WSE. These are firms controlled by the Polish state and for most of them 100% of the shares are held by the state.
In contrast to the privatization programs in Russia, the Czech Republic and Slovakia, an MPP was carried out in Poland only in a second phase of the privatization program. At the end of 1994, within the framework of the program, 15 National Investment Funds (NIFs) were established in the form of joint-stock companies. 512 small- and medium-sized state enterprises took part in the MPP. In March 1995, each fund took a majority holding (33%) in 33 to 34 enterprises and a minority holding in all other companies participating in the MPP. In this way, the 15 NIFs held a total of 60% of the shares of each company. Up to 15% of the shares of each participating company were made available free of charge to its employees, and 25% were retained by the State Treasury.4
In the MPP each adult Pole was entitled to receive one National Investment Funds (NIF) certificate for a fee of 20 PLN, which could later be converted into shares of the 15 NIFs (one share of each of the 15 funds for one certificate). The certificates were dis-tributed between November 22nd, 1995 and November 22nd, 1996. The WSE began listing the certificate on July 15th 1996. The conversion process of one certificate into 15 NIF shares began on May 12th, 1997 and lasted until the end of 1998. Trading in the 15 NIF shares began on June 12th, 1997.5
In total about 25.55 NIF certificates were converted into NIF shares. More than 91% of the 28 million citizens who were eligible for the MPP took part. Since the certificate was officially traded on the WSE Polish citizens had the opportunity to decide whether to sell the certificate on the aftermarket or to convert them into NIF shares.
To compare the price behavior of an investment in the MPP with those of the case-by-case privatization IPOs, a separate MPP sample is used. It starts on July 15th 1996 with the listing start of the certificate on the WSE. Until June 11th, 1997 the market prices of the certificate and from June 12th, 1997 until July 14th, 1999 the market prices of the 15
4 See WSE Fact Book (1999) and Dietrich et al. (1996).
5
NIF shares are used. The price history of the MPP sample therefore comprises three years, which enables a comparison with the three-year long-run aftermarket perform-ance of the two IPO samples.
Table 1 presents descriptive statistics of all samples. The sample All Issues (All) consists of 149 firms, the sample Privatization IPOs (PIPOs) of 51 firms and the sample Private
Sector IPOs (IPOs) of 98 firms. First, about 85.5% of the total gross proceeds of
15,874.73 million PLN comes from the 51 PIPOs (13,498.27 Million PLN). The mean gross proceeds for the PIPO firms (264.67 million PLN) are more than 11 times higher than for the private sector IPOs (23.33 million PLN; median values: 42 million PLN versus 14.03 million PLN). The mean as well as the median values are significantly different from each other (see Table 2). This has two reasons: (i) First, PIPO firms are larger than private sector IPOs. (ii) Second, the fraction sold at the initial offer is more than twice as high for the PIPO firms (63.9%) as the corresponding value for the private sector IPOs (average value 32.22%). Table 2 shows that the two samples significantly differ in this respect from each other. This applies to the mean as well as the median.
Privatization IPOs and private sector IPOs not only differ with respect to the gross pro-ceeds but also significantly depending on whether secondary or primary shares are sold at the initial offer. 93.4% of the gross proceeds for all PIPO firms consist of secondary shares and for two thirds of all privatization IPOs only secondary shares are sold.
The high fraction of the capital sold at the initial offer and the huge portion of secondary shares can be viewed as a first indication that the Polish government is market-oriented in the sense of Perotti (1995). A big portion of secondary shares is an indication that the government is trying to privatize fast. Issuing only primary shares, a large reduction of the fraction held by the state can only be reached through a huge capital increase. This has the disadvantage that (i) the state treasury would receive no money and (ii) the firm suddenly gets an extreme amount of money, which it could perhaps not use properly. By privatizing through selling secondary shares the managers and the new owners can de-cide on their own when and to what extent they will raise new capital.
In contrast to privatization IPOs only primary shares are sold at the initial offer in 84.7% of all private sector IPOs. Their going public process obviously serves to in-crease the capital base.
Biais and Perotti (1997) show that, if there is great uncertainty about the future privati-zation policy, a market-oriented government having difficulties in signaling its identity can privatize successfully by additionally using fixed-price offers (firm commitment offerings). As Benveniste and Busaba (1997) note, this can generate a higher demand for an issue. The empirical evidence shows first that only 59% of all privatization IPOs in the sample are equipped with a fixed issue price and, second, that the average de-mand measured by the dede-mand multiple for the privatization IPOs is only a little higher than for the private sector IPOs (2.75 versus 2.07). The mean and median demand mul-tiple values of the two samples are not significantly different from each other (see Table 2).6
Table 1 and 2 also show that the time period from the start of the subscription period till the first trading day is significantly lower for privatization IPOs than for private sector IPOs (61 days vs. 101 days). Assuming that the risk of an issue increases with this time period and that investors demand a premium to take this risk, it is sensible for a gov-ernment to decrease this period to be able to get a higher issue price.
Biais and Perotti (1997) predict that governments allocate shares in a political inspired manner. To achive political goals governments divide issues in several tranches and prefer e.g. employees and domestic retail investors when assigning a certain number of shares to each tranche. In Poland trade unions have historically been very powerful. It can therefore be expected that the government will particularly try to participate the employees of the state enterprises on offer. On average (median) 19.4% (18.7%) of the shares sold during the initial offer are sold to employees. This value is more than twice as high as the international evidence in Jones et al. (1999). For a 59-country sample they report an average (median) portion sold to employees of 8.5% (7.0%). This underlines
6 In a tender offer the issue price is not fixed before the start of the subscription period. Only a
the strong and important position of trade unions in Poland and the political necessity to integrate Polish employees in the privatization program to a higher degree than in most other countries.
3. Initial Return
3.1 Buy-and-Hold Return and Buy-and-Hold-Abnormal Return
The starting point to investigate the initial returns is the calculation of buy-and-hold returns (BHRs) as well as buy-and-hold-abnormal returns (BHARs). The BHR for IPO i corresponds to the initial raw return and is defined as
0 , i 0 , i 1 , i i P P P BHR = − , (1)
Pi,0 represents the issue price and Pi,1 the closing price on the first trading day of IPO i.
The time index 0 refers to the first day of the subscription period. To measure the initial excess return or buy-and-hold abnormal return (=underpricing) the market-adjusted re-turn method is used. The initial excess rere-turn for each IPO i is then defined as the dif-ference between its initial raw return (the BHRi) and the corresponding return on the
market index: i i , WIG i i BHR BHR UP BHAR = − = (2)
where BHARi is the buy-and-hold abnormal return of IPO i, BHRWIG,i is the
buy-and-hold return of the Warsaw Stock Exchange Index (WIG) and UPi is the underpricing of
IPOi.7 Analogous to (1) the BHR for the WIG-Index is defined as
7
0 , i 0 , i 1 , i i , WIG WIG WIG WIG BHR = − . (3)
WIGi,0 represents the WIG-Index on the first day of the subscription period of IPO i and
WIGi,1 is the WIG-Index at the end of the first trading day of IPO i. In addition wealth
relatives (WR) are calculated. In accordance with Ritter (1991) the WR of IPO i (WRi)
is defined as i , WIG i i BHR 1 BHR 1 WR + + = . (4)
Table 3 shows BHRs (initial raw returns), BHARs (underpricing) and wealth relatives for all samples. The average underpricing of Polish IPOs (sample All Issues) is +35.57% and is significantly different from zero. The same applies to the median (+15.14%). About a quarter of all Polish IPOs are overpriced (BHAR < 0). The null hypothesis of no (average) underpricing has to be rejected. With the trading strategy of investing a similar amount of money in each of the 149 Polish IPOs in the sample All Issues at the issue price and selling each IPO on its first trading day an investor would have earned an initial raw return of +41.29% (not allocation adjusted).
The BHRs and the underpricing of the Privatization and the Private Sector IPO sample are significantly different from zero. The empirical evidence in other studies concerning the underpricing of privatization and private sector IPOs is not unambiguous. There is evidence for a higher underpricing of privatization IPOs (see e.g. Menyah and Paudyal (1996) for the UK or Paudyal (1998) for Malaysia) as well as evidence for no signifi-cant difference between the two groups (see e.g. Dewenter and Malatesta (1997) for 8 countries or Jones et al. (1999) for 59 countries).
From the model of Perotti (1995) it could be concluded that for a market-oriented gov-ernment the underpricing of privatization IPOs should be higher than for private sector IPOs. The reason for this is that in order to build up reputation and to signal its identity a market-oriented government sells state enterprises for issue prices which are below the
prices owners other than the state would demand and receive. If the price on the first trading day corresponds to the firm value, the initial return for privatization IPOs should therefore be higher. This hypothesis is also in line with the observation that the PE-ratio of privatization IPOs is significantly lower than for private sector IPOs.
A comparison of privatization IPOs and private sector IPOs shows that on average pri-vatization IPOs have an underpricing which is about 40 percentage points above those of private sector IPOs. But this huge difference in the mean is not significantly different from zero at conventional significance levels (see Table 4, Panel A). Only the differ-ence in the underpricing-median is significant at the 10%-level. The main reason for the large mean underpricing of privatization IPOs is one issue: Bank Slaski. After an issue price of 50 (new) PLN the market price on the first trading day rose to 675 (new) PLN. The resulting BHR of +1,250% caused a lot of criticism for the privatization policy and also political consequences. Without Bank Slaski the underpricing of the privatization IPOs drops to +39.96% (Median: +18.01%, Wealth Relative: 1.384), but is also signifi-cantly different from zero at the 1%-level.
The average underpricing of Polish privatization IPOs is a little higher than the +34.1% (Median: +12.4%) reported in Jones et al. (1999) for the 59 country sample of 303 share issue privatizations. A comparison with the initial return of the MPP shows a hardly surprising result: The issue price in the MPP was fixed as a fee at a very low level of 20 PLN to guarantee that a lot of Polish citizens take part in the program. The closing price of the certificate on the first trading day was 104 PLN. Each Polish citizen taking part in the MPP achieved an immediate profit of 84 PLN or +420%.
The model of Perotti (1995) predicts that in an environment of high uncertainty about the future privatization policy, issues of which the government sells a higher fraction at the initial offer should experience a higher underpricing. The empirical evidence in Ta-ble 4, Panel B, shows that this applies for the mean initial return of Polish privatization IPOs. The differences in the mean BHR, BHAR and WR are all significantly different from zero at the 10%-level. The median differences are all positive but not significant. The distribution of the differences is therefore strongly skewed to the right. No
signifi-cant differences can be observed in the mean and median values for the private sector IPOs.
The uncertainty about the future privatization policy is expected to be highest at the start of a privatization process. This implies that the fraction sold at the initial offer and the underpricing should be highest at the start of the privatization program. A market-oriented government should build up reputation regarding its privatization policy in the course of time. If this applies to Poland the fraction sold at the initial offer and the un-derpricing should decrease over time. To analyze these hypotheses the investigation period is divided into four subperiods: 1991 till 1992 (subperiod one), 1993 till 1994 (subperiod two), 1995 till 1996 (subperiod three) and finally 1997 till 1998 (subperiod four).
Table 5 reports the empirical results. First, for PIPOs the fraction sold from the capital at the initial offer decreases in the course of time (see Table 5, Panel B). The average (median) fraction sold is in the first subperiod 86.1% (90.0%), 63.6% (68.3%) in the second subperiod, 54.6% (60.1%) in the third subperiod and 49.3% (47.4%) in the last subperiod. Second, the underpricing level of PIPOs is not highest in the first but in the second subperiod and decreases thereafter (see Table 5, Panel A). The average (median) underpricing for PIPOs is 11.8% (13.1%) in the first subperiod, 191.0% (71.4%) in the second subperiod, 46.8% (29.2%) in the third subperiod and 13.8% (14.8%) in the last subperiod. The low underpricing and the high fraction sold in the first subperiod would suggest a populist government. A populist government tries to maximize gross proceeds and therefore sells higher fractions at the initial offer at not too small issue prices, re-sulting in a lower underpricing. But another interpretation is that the first subperiod has been used as a test phase by the Polish government. This view is supported by the fact that the price-earnings (PE) ratios (and therefore the issue prices) are very low in the first subperiod. In addition, state-owned enterprises sold in the first subperiod are very small compared to PIPOs in the remaining periods. The average (median) gross pro-ceeds of PIPOs in the second to fourth subperiod are about 19 (3) times larger than in the first subperiod.
A further analysis of the PE-ratios additionally supports the view that the Polish gov-ernment is market-oriented. In all subperiods but the last the average and median PE-ratios of PIPOs are below those of private sector IPOs (see Table 5, Panel C). A populist government would not sell state enterprises over such a long time at prices which are consistently below those of private sector IPOs. In accordance with the model of Perotti (1995) this results indicate that the Polish government deliberately sold state enterprises at low prices to build up reputation and to signal its identity as market-oriented. This evidence is in addition confirmed by the fact that for private sector IPOs the underpric-ing levels, the fractions sold at the initial offer and the PE-ratios do not behave similar compared to PIPOs. In the course of time the fraction sold is nearly constant at around 30%, the underpricing level decreases, but on a much lower level than for PIPOs and the mean as well as median PE-ratios do not vary much. In the last period (1997 till 1998) the underpricing level and the PE-ratios of PIPOs and private sector IPOs are nearly similar. Only the mean and median fraction sold at the initial offer are higher for PIPOs (49.3% and 47.4%, respectively) compared to private sector IPOs (30.3% and 30.0%, respectively). This is an indication that the process of building up reputation has nearly reached an end in Poland.
3.2 Explanations for the level of underpricing
To examine the level of underpricing for Polish IPOs, the most common explanation models are tested. The samples All Issues, Privatization IPOs and Private Sector IPOs are investigated separately. This allows getting additional insight into the differences between privatization IPOs and private sector IPOs.
(a) Winner’s Curse
In Rock’s model (1986) underpricing is a necessary equilibrium condition in a world of informational asymmetry between groups of informed and uninformed investors. An informed investor can distinguish between underpriced and overpriced new issues and
will therefore submit purchase orders only if the offer price is less than the true value of the stock. If, on the other hand, the issue is overpriced, only uninformed investors sub-mit purchase orders and they subsequently receive a full allocation.
For underpriced issues, both uninformed and informed investors submit purchase orders and the allocation is subsequently rationed between the two groups. This creates a situation where the average first-day return conditional upon receiving shares is lower than the average first-day return conditional upon submitting a purchase order. There-fore, uninformed investors face a "winner's curse": The chance of being allocated shares in overpriced new issues is greater than in underpriced issues. To keep uninformed in-vestors participating in the IPO market, issues are underpriced to ensure them a non-negative, market-adjusted rate of return. To test whether Rock's (1986) winner's curse hypothesis holds for the sample under investigation, it is necessary to calculate alloca-tion-adjusted abnormal returns. This requires information on the amount of rationing.
In contrast to most other countries, the allocation rates are published in Poland. For firm commitment offerings the rationing is done using a uniform allocation rate. Uniform means that for oversubscribed issues all orders are rationed proportionally. If for par-ticular groups of investors (e.g. retail investors, institutional investors, employees, for-eign investors) a particular amount of shares is reserved in advance, each group of in-vestors has their own allocation rate. Therefore for firm commitment offerings an exact allocation-adjusted initial return can be calculated. For tender offers there are two pos-sibilities: (i) a uniform allocation rate for all orders or (ii) an allocation depending on the submitted price. To calculate allocation-adjusted initial returns for tender offers in (ii) an average allocation rate is used as an approximation.
The results in Table 6 show that the allocation-adjusted BHARs (allocation-adjusted underpricing) is much lower than the non-adjusted values (see Table 3). This can be explained by the fact that first a low issue price and therefore a higher underpricing generates a higher demand for an issue and that second a higher demand leads to a lower allocation rate. The correlation between the demand multiple and the underpricing
is +0.39 and between the demand multiple and the allocation rate is -0.80. Both values are significantly different from zero at the 1%-level.
The allocation-adjusted BHARs are also significantly different from zero for all sam-ples. Therefore the underpricing of Polish IPOs can only partly be explained by the winner’s curse. In the multiple regression analysis below the demand multiple is used as an explanatory variable. The results presented above suggest that a positive relationship between the underpricing level and the demand multiple can be expected.
(b) Ex-ante uncertainty about the firm value
Another testable implication of Rock’s model (1986) is that the level of underpricing required to attract uninformed investors into the market increases with the ex-ante un-certainty about the true value of the firm. The greater the ex-ante unun-certainty, the greater is the advantage of becoming an informed investor. Beatty and Ritter (1986) suggest that there should be a positive relationship between the underpricing and the non-observable ex-ante uncertainty. In order to test their proposition it is necessary to use a proxy for the ex-ante uncertainty. The standard deviation of daily returns during the first two trading months of aftermarket trading (exclusive of the initial return) is used as a proxy in the multiple regression analysis. According to Beatty and Ritter’s model (1986) a positive relationship can be expected between the underpricing and the ex-ante uncertainty.
For another test whether asymmetric information between managers/owners and inves-tors affects the underpricing level the firm size can be used. The larger the firm on offer, the more the firm should be known to potential investors and the lower the underpricing should be. Therefore the market value of the issues (on the first trading day) is used as a proxy for firm size as an explanatory variable in the multiple regression analysis below. If asymmetric information about the firm value leads to a higher underpricing, the rela-tionship between underpricing and size should be negative.
(c) Signaling Hypothesis
Studies by Allen and Faulhaber (1989), Grinblatt and Hwang (1989) and Welch (1989) suggest that underpricing may itself be a costly signal of the intrinsic value of the issu-ing firm. In the Welch (1989) signalissu-ing approach for example, high-quality firms delib-erately choose an offer price below the intrinsic value to signal their quality to investors. This underpricing is motivated by the possibility of achieving higher offer prices in sub-sequent seasoned issues. Underpricing is therefore supposed to be a sort of "appetizer" for following issues in the aftermarket.
High-quality firms (= high underpricing) should therefore only offer a smaller part of their share capital at the initial issue and make subsequent issues in the aftermarket. If this explanation model applies, the level of underpricing should be proportional to the frequency of seasoned issues and inversely proportional to the percentage fraction sold from the share capital. To test this proposition the fraction sold at the initial offer and a dummy variable for the seasoned equity offerings (SEOs) is used in the multiple regres-sion analysis.
(d) Political uncertainty
The model of Perotti (1995) predicts that if the government is market-oriented the rela-tionship between to underprice level and the fraction sold at the initial offer should be positive. This is in contrast to the prediction of a negative relationship by the traditional IPO signaling models.
According to Biais and Perotti (1997) a market-oriented government can generate sup-port for its privatization program by allocating more shares to median class investors. It can achieve this aim on a large scale by selling a larger portion at the initial offer. Therefore the underpricing should be higher for those issues of which a relatively large portion is sold at the initial offer and of which a relatively large portion is allocated to domestic retail investors. This higher underpricing is necessary to generate demand from this particular group of investors.
3.3 Multiple Regression Model
To test the above models and hypotheses the following regression models are tested:
Model I: i i 5 i 4 i 3 i 2 i 1 o
i DM Size Risk SEO LF
UP =α + α ⋅ + α ⋅ + α ⋅ +α ⋅ + α ⋅ +ε Model II: i i 5 i 4 i 3 i 2 i 1 o
i DM Size Risk SEO LFS
UP =α + α ⋅ + α ⋅ + α ⋅ +α ⋅ + α ⋅ +ε
UPi is the underpricing, DMi the demand multiple, Sizei the logarithmic market value
and Riski the standard deviation of daily returns over the first two months in the
after-market (proxy for the ex-ante uncertainty), for IPO i. SEOi, LFi and LFSi are dummy
variables. SEOi is coded one if a seasoned offering is made within 3 years of the IPO
and zero otherwise; LFi is coded one if a large fraction of IPO i is sold (i.e. a fraction
above the median fraction) and zero otherwise; LFSi is coded one if a large fraction of
IPO i is sold and if the portion of the capital owned by domestic retail investors after the issue is at least 25% (and zero otherwise). Thus, in model II an additional investigation is made to find out whether it is important for the underpricing level to whom a gov-ernment sells shares if it sells a large fraction.
3.4 Regression Results
Table 7 shows the following regression results for models I and II: First, the demand multiple (DM) is significantly positively related to the underpricing level. This applies to the sample All Issues as well as the two subsamples Privatization IPOs and Private Sector IPOs. This result is evidence that part of the underpricing can be explained by the winner’s curse model. The better an issue is rated based on the relationship between issue price and market value, the higher is the underpricing and the more orders are placed by informed investors, which results in a higher demand. Second, the ex-ante uncertainty (at least the used proxy) has no explanatory power. This applies to privati-zation IPOs as well as private sector IPOs.
Third, the underpricing level of privatization IPOs is significantly positively related to the firm size. The larger the firm the higher the underpricing. According to the explana-tion hypothesis of an asymmetric informaexplana-tion distribuexplana-tion between managers/owners and investors about the firm value, smaller firms should have a high underpricing be-cause they are less known. But exactly the opposite is true of privatization IPOs. This supports the hypothesis that the Polish government sells big and well-known state en-terprises at a lower issue prices to build up reputation. This enables the government to generate support for its privatization policy. This interpretation is also supported by the fact that the variable size is not in a significantly positive relationship with the under-pricing level for private sector IPOs.
Fourth, underpricing is for privatization IPOs no signal for a „good“ firm in the context of the traditional signaling hypotheses. No significantly positive relationship between the underpricing level and the SEO activity is observable. This is in contrast to the evi-dence for private sector IPOs. For them, the relationship between underpricing and SEO activity is significant (at the 10%-level). Private sector IPOs increasing their share capital by making an SEO experience a higher underpricing.
Fifth, the hypothesis that for privatization IPOs the underpricing increases with the fraction sold at the initial offer can be supported (significant at the 10%-level). The un-derpricing is the higher, the higher the fraction sold. These observations are in line with the model of Perotti (1995). In contrast to privatization IPOs there is no significant rela-tionship between fraction sold and underpricing observable for private sector IPOs.
Sixth, model II shows that the underpricing for privatization IPOs is especially large if a large fraction is sold and if in addition domestic retail investor receive a large portion at the initial offer. If the governments wants to increase the number of shareholders and therefore also the acceptance of the privatization policy among the citizens, it uses a low issue price, which lead to a higher underpricing. This empirical evidence is in accor-dance with the model of Biais and Perotti (1997).
4. Aftermarket Performance
The aim of this section is to examine the aftermarket performance of Polish IPOs and the MPP. Besides the short-run aftermarket performance, which permits conclusions about the price adjustment process, the long-run performance within three years of the first trading day is analyzed. The latter is of multiple interest: First, the long-run per-formance of privatization IPOs is important for a market-oriented government that is trying to build up reputation and support for its privatization policy. It should be inter-ested in making sure that the long-run performance of privatization IPOs is not negative. Second, the same applies to the long run price behavior of the MPP the Polish govern-ment carried out in the mid nineties.
Third, for private sector IPOs often a negative long-run performance is documented (see e.g. Loughran, Ritter and Rydqvist (1994)). What has not been documented in the lit-erature yet is whether private sector IPOs in a transition economy in which the govern-ment has been carrying out a privatization program for several years behave in the same manner as private sector IPOs in developed markets or whether significant differences are observable. Fourth, in a transition economy the restructuring of firms plays an im-portant role. The question arises as to whether one aspect of the restructuring efforts, the raising of new capital, has a measurable influence on the long-run performance and, if so, in which direction.
4.1 Methodology
To measure the performance of IPOs in the aftermarket buy-and-hold returns are calcu-lated in a first step for each IPO i. In contrast to cumulative returns (which are some-times used to measure long-horizon security price performance) buy-and-hold returns have the advantage that they are based on a realistic ex-ante trading strategy. The buy-and-hold return for IPO i (BHRi,T) is defined as
1 R 1 BHR T 1 t i,t T , i − ∏ + = = (5)
where Ri t, is the return of IPO i in period t and t=1 indicates the first trading day in the aftermarket.
BHRs are calculated for the following time periods: T = 1 week, 2 weeks, 1 month, 2 months, 1 year, 2 years and 3 years. With this the short- and long-run price behavior of Polish IPOs on the aftermarket can be investigated. During the investigation period four delistings occurred: three mergers and one going private. The performance measure-ment for the first three years of aftermarket trading has to take into account only one merger (Bank Gdansk). The three other delistings are outside the first three years. For Bank Gdansk the returns of the absorbing firm (BIG) are used from the delisting till the third anniversary.
The aftermarket performance is measured against the WIG-Index as a benchmark. The use of market indices as benchmarks is common in the literature (see e.g. Keloharju (1993) for the Finnish IPO market, Kunz and Aggarwal (1994) for the Swiss IPOs mar-ket or Paudayal (1998) for the IPO marmar-ket in Malaysia). A second possibility is to use a matching firm adjustment procedure (cf. e.g. Ritter (1991) or Loughran and Ritter (1995)). This requires a sufficient number of potential matching firms (Non IPOs). But this is not the case for Poland.
In a similar way to (5) the BHR of the WIG-Index for IPO i (BHRi, WIG, t) is calculated
as 1 R 1 BHR T 1 t i,WIG,t T , WIG , i − ∏ + = = . (6)
Ri,WIG,t is the return of the WIG-Index in period t, where t=1 indicates the first trading
day in the aftermarket. Therefore, buy-and-hold returns over identical intervals are cal-culated for each IPO and its corresponding reference portfolio. To measure the
market-adjusted performance buy-and-hold abnormal returns (BHARs) and wealth relatives (WRs) are used.
Another important point in measuring the long-horizon abnormal performance of secu-rity prices is the usage of appropriate test statistics. As the simulation results of Kothari and Warner (1997), Barbar and Lyon (1997) and Lyon, Barbar and Tsai (1999) show, conventional tests of long-run abnormal security returns are often misspecified. They find that conventional parametric test statistics often indicate a long-run abnormal per-formance when none is present. They mention especially three main reasons for poten-tial misspecifications: (a) survival-related biases, which occur if failing firms are ex-cluded, (b) rebalancing biases, which arise if cumulative return procedures are used and (c) biases because long-run abnormal performance measures are typically skewed.
To minimize these sources of misspecification, this study also includes all firms delisted on the WSE during the investigation period and uses buy-and-hold returns to calculate the long-run performance. In addition, to account for the skewness bias, a skewness-adjusted t-statistic with bootstrapped p-values (as suggested by Lyon, Barbar and Tsai (1999)) and a non parametric Wilcoxon-Signed-Rank-Test are used to test the null hy-pothesis of no abnormal long-run performance.8
4.2 Results
To be able to compare the aftermarket performance within the first three years after the first trading day for different time periods (e.g. one year and three years) only issues are used with a first trading day earlier than January 1st, 1996. This leads to a reduction of the sample All Issues to 57 firms, the sample Privatization IPOs to 33 firms and the sample Private Sector IPOs to 24 firms.
8 As a result of their simulation analyses concerning the long-horizon security price performance,
Kothari and Warner (1997) recommend using nonparametric and bootstrap tests to reduce misspecification.
To examine the aftermarket performance of the MPP for the first three years market prices of the National Investment Funds certificate are used from 15.7.1996 till 11.6.1997 and the market prices of the 15 National Investment Funds (NIFs) shares from 12.6.1997 till 14.7.1999.
Panels A and B of Table 8 show the aftermarket performance for the different samples. First, it can be seen that in line with the evidence of many other markets the short-run aftermarket performance is not significantly different from zero. For the sample All Is-sues e.g. the average BHAR over the first two month is -4.47% (Median: -6.16%). A slight negative short-run aftermarket performance can also be observed for privatization IPOs and the private sector IPOs. All mean and median values are not significantly dif-ferent from zero. This result leads to the conclusion that for Polish IPOs there is full price adjustment on the first trading day.
The long-run performance for the first three years shows clear differences between the samples. The sample All Issues does not experience a significantly negative or signifi-cantly positive performance (mean-BHAR: +20.09%; median-BHAR: -45.42%, WR: 1.052). The same applies to privatization IPOs: the mean-BHAR is relatively high and positive (+61.45%), but the median-BHAR is negative (-23.67%). Both values are not significantly different from zero at conventional significance levels.
This evidence for the privatization IPOs is in contrast to the international evidence found by Megginson et al. (1998). Using country indices as benchmarks, they document for a 36-country sample of 134 privatization IPOs a significantly positive aftermarket performance for the first three years of aftermarket trading (mean-BHAR: +54.9%, me-dian-BHAR: +26.0%). But it has to be noted that this result is primarily caused by the very strong performance of UK privatization IPOs. If UK privatization IPOs are ex-cluded from the sample, the Mean-BHAR decreases to +30.1% and the Median value drops to -6.3%. For non-UK privatization IPOs the long-run performance is therefore no longer unambiguous.
In contrast to the evidence for privatization IPOs private sector IPOs experience a sig-nificantly negative performance over the first three years (see Table 8, Panel A and B). The mean-BHAR (-36.77%) as well as the median-BHAR (-59.11%) are significantly different from zero at the 10%-level. Private sector IPOs in Poland therefore behave in a similar manner as private sector IPOs of other markets do: they significantly underper-form in the long-run.9
These results show that privatization IPOs and private sector IPOs differ from each other with respect to the long-run price behavior. According to the model of Perotti (1995) it could be expected that a market-oriented government trying to build up repu-tation over time is not interested in a significantly negative long-run performance of privatization IPOs. The non-negative long-run performance of the Polish privatization IPOs can be viewed as additional evidence that the Polish government is market-oriented.
The evidence of the long-run performance for the MPP is surprising. In strong contrast to the evidence for the privatization IPOs (case-by-case privatization) and the expecta-tions in Perotti (1995) the long-run aftermarket performance of the MPP within the first three years after the first trading day of the NIF certificate is significantly negative at the 1%-level: mean-BHAR = -69.41%, median-BHAR: -74.22% (see Table 8). The same applies to the BHRs: mean-BHR = -39.9%, median-BHR: -44.7%. An investor buying a NIF certificate on its first trading day and converting the certificate into 15 NIF shares thereafter lost 39.9% of his initial investment over the fist three years!
Compared to the volume of case-by-case privatizations in Poland the MPP is of less importance. This gives rise to the hypothesis that the MPP was initiated by the govern-ment mainly as a device to intensify the participation of Polish citizens in the privatiza-tion process. The very low issue price is an indicaprivatiza-tion that the government tried to gen-erate a positive mood for its privatization policy among the citizens. But the negative aftermarket performance of the MPP opposes this aim.
9 See e.g. the evidence for US-IPOs in Loughran and Ritter (1995) or for finnish IPOs in Keloharju
Now the following question arises: What is the total performance over three years for a Polish citizen buying a NIF certificate at the issue price of 20 PLN, converting it into shares of the 15 NIFs and selling the 15 NIF shares three years after the first trading day? This question is of importance for a market-oriented government because if the originally very high initial return disappears over time, a possibly positive political ef-fect of the MPP would only be small or even negative. Table 8 shows the performance results including the allocation-adjusted initial return in panels C and D.
A Polish citizen experiences a significantly positive BHR of +212.7% (BHAR: +95.24%) from the issue price of 20 PLN till the third anniversary of public trading. Because of the small issue price the MPP provides (despite the negative aftermarket performance) a significantly better performance than an alternative investment in the Polish Stock Market (WIG-Index). But it remains unclear why the MPP provides a sig-nificantly negative aftermarket performance in relation to the case-by-case privatiza-tions. The next one to two years will show whether the negative performance trend of the MPP lasts or whether the performance of the MPP (inclusive of initial return) will collapse with those of the case-by-case privatization program.
4.3 Restructuring Hypothesis
As for IPOs there is also evidence of a significantly negative long-run performance for SEOs (see e.g. Loughran and Ritter (1995) for the US or Cai and Loughran (1998) for Japan). In contrast to this evidence Megginson et al. (1998) document a non-negative (zero) long-run performance for their 36-country sample of privatization IPOs. In tran-sition economies changing to a market-oriented system most firms have a huge restruc-turing need. This requires not only a lot of technical and personal know-how, but also a lot of additional capital. This leads to the (restructuring) hypothesis that for firms ab-sorbing new capital the restructuring process proceeds better and faster. If this hypothe-sis applies firms in an emerging market like Poland taking up new capital should pro-vide a better long run performance than firms with no additional capital.
Table 9 shows the results of an analysis of the differences in the aftermarket perform-ance between Polish IPOs making a SEO within the first three years and those without an SEO (NoSEO). In accordance with the restructuring hypothesis the SEO-firms expe-rience a significantly better long-run performance than the NoSEO firms. The differ-ences in the mean-BHAR of +180.35 percentage points is significantly different from zero at the 5%-level.
5. Summary
This study examines the going public process in Poland. In contrast to other papers the characteristics and the short-and long-run price behavior of three groups of firms are compared with each other: (i) Case-by-case privatization initial public offerings (PIPOs), (ii) private sector initial public offerings (IPOs) and (iii) certificates and shares of the mass privatization program (MPP). Unlike other Central and Eastern European countries (e.g. Russia, Czech Republic or Slovakia) Poland did not start the privatiza-tion process with an MPP; instead case-by-case privatizaprivatiza-tions were carried out over the first years of the transition. An MPP was started in the mid nineties with the distribution of National Investment Fund (NIF) certificates to Polish citizens, who after a while were able to convert the certificates into shares of 15 National Investment Funds (NIF shares). As the certificates were and the NIF shares still are officially traded on the WSE, this provides the rare opportunity to compare case-by-case privatizations and the MPP on the basis of their long-run performance.
The following questions are examined:
(i) Can the Polish government in the sense of Perotti (1995) be classified as market-oriented and is a reduction in the political uncertainty about the future privatiza-tion policy observable in the course of time?
(ii) How do case-by-case privatizations, private sector IPOs and the MPP differ with respect to underpricing? Is the underpricing of case-by-case privatizations and the MPP higher than those of private sector IPOs?
(iii) How can the underpricing of case-by-case privatizations and private sector IPOs be explained? Are explanation hypotheses originally designed for developed markets also valid for Polish private sector IPOs and to what extent do they ap-ply to do case-by-case privatizations?
(iv) Are there differences in the long-run performance between case-by-case privati-zations, private sector IPOs and the MPP? Which group has the best perform-ance?
(v) In a transition economy, companies have an increased restructuring need. Is the raising of new capital an indication for a faster restructuring process and can this be seen in the long-run performance?
The results of the investigations are as follows:
The comparison between case-by-case privatizations and private sector IPOs clearly shows that in accordance with Perotti (1995) the Polish government can be classified as market-oriented. At the start of the privatization process the uncertainty over the future privatization policy is highest. In that phase the Polish government sold a larger fraction of state owned enterprises during the initial offer. During the period under investigation (April 1991 till December 1998) this fraction dropped, which can be interpreted as an indication for a build-up of reputation. Compared to private sector IPOs case-by-case privatizations are sold at lower prices (lower PE-ratios). This applies especially to the first part of the privatization process. By contrast, a populist government would not sell state enterprises over a long period of time at prices that, on average, are below those of private sector IPOs.
Case-by-case privatizations as well as private sector IPOs show on average a significant underpricing. With a mean of +62% the underpricing of privatization IPOs is about 40 percentage points above the underpricing of private sector IPOs. But only the median
value of the underpricing difference is significantly different from zero (at the 10%-level). Because of a very low issue price the underpricing of the MPP is +352%. About 91% of the eligible Polish citizens collected their NIF certificates. Many people made use of the high initial return and sold their certificates shortly after collecting them.
The underpricing of case-by-case privatizations and private sector IPOs is caused to a large extent by the winner’s curse. The allocation-adjusted underpricing drops by about 75% for privatization IPOs and 50% for private sector IPOs. With increased underpric-ing the demand for shares at the initial offer increases and a higher rationunderpric-ing follows. In accordance with Perotti (1995) the underpricing is higher if a higher fraction of the capital is sold at the initial offer. In contrast to that, the asymmetric information between managers/owners and investors about the firm value as well as the ex-ante uncertainty about the firm value have no explanatory power for the underpricing level of privatiza-tion IPOs. The same applies to signaling models.
The long-run performance shows interesting results. In contrast to Megginson et al. (1998), who examine the (average) long-run performance for a 36-country sample, Pol-ish case-by-case privatizations do not show a significantly positive performance over the first three years of aftermarket trading. The long-run performance of private sector IPOs is significantly negative, which is similar to the evidence for many developed markets (see e.g. Loughran, Ritter and Rydqvist (1994)). The non negative long-run performance of case-by-case privatizations is consistent with a market-oriented gov-ernment trying to build up reputation for the privatization program in the course of time. What is surprising is the significantly negative long-run performance of the MPP. After three years of aftermarket trading, its wealth relative (WR) of 0.495 is even below the WR for private sector IPOs (WR=0.802). One would not expect this for a market-oriented government. When the initial return is included in the long-run performance, the long-run performance for the MPP is significantly positive (BHR=+213%; BHAR=+95%). Therefore Polish citizens who did not sell their MPP certificates during the early stages of aftermarket trading did not lose money on their initial investment. The next one to two years will show whether the negative performance tendency of the
MPP lasts or whether the overall performance of the MPP collapses with those of the case-by-case privatization program.
Finally it is documented that Polish IPOs making an SEO during the first three years after the first trading day experience a significantly better long-run performance (BHAR=+180 percentage points) than IPOs without SEOs. This evidence is in contrast to the evidence of a negative long-run performance of SEO firms in developed markets but supports the restructuring hypothesis. Accordingly, firms in a transition economy trying to restructure fast should perform better in the long-term. SEOs can therefore be an indication of increased restructuring activities in a transition economy.
References
Allen, F. and G.R. Faulhaber (1989), Signaling by Underpricing in the IPO Market, Journal of Financial Economics, Vol. 23, 1989, pp. 303-323.
Barber, B.M. and J.D. Lyon (1997), Detecting long-run abnormal stock returns: The empirical power and specification of test statistics, Journal of Financial Eco-nomics, Vol. 43, 1997, pp. 341-372.
Beatty, R.P. and J.R. Ritter (1986), Investment Banking, Reputation, and the Underpricing of Initial Public Offerings, Journal of Financial Economics, Vol. 15, 1986, pp. 213-232.
Benveniste, L.M. and W.Y. Busaba (1997), Bookbuilding vs. Fixed Price: An Analy-sis of Competing Strategies for Marketing IPOs, Journal of Financial and Quantitative Analysis, Vol. 32, No. 4, December 1997, pp. 383-403.
Biais, B. and E.C. Perotti (1997), Machiavellian Underpricing, Working Paper, Uni-versite des Sciences Sociales de Toulouse, University of Amsterdam and CEPR, December 1997.
Boubakri, N. and J.-C. Cosset (1998), The Financial and Operating Performance of Newly Privatized Firms: Evidence from Developing Countries, Journal of Fi-nance, Vol. 53, No. 3, pp. 1081-1110.
Cai, J. and T. Loughran (1998), The performance of Japanese seasoned equity offerings, 1971-1992, Pacific-Basin Finance Journal, Vol. 6, pp. 395-425. Carter, R.B. and S. Manaster (1990), Initial Public Offerings and Underwriter
Reputation, Journal of Finance, Vol. 45, No. 4, Sep. 1990, pp. 1045-1067. D’Souza, J. and W.L. Megginson (1999), The Financial and Operating Performance of
Privatized Firms During the 1990s, Journal of Finance, Vol. 54, No. 4, Aug. 1999, pp. 1397-1438.
Dewenter, K.L. and P.H. Malatesta (1997), Public Offerings of State-Owned And Privately-Ownes Enterprisies: An International Comparison, Journal of Finance, Vol. 52, No. 4, pp. 1659-1679.
Dietrich, M., D. Ipsen and R. Puntillo (1996), Poland’s mass privatization program, European Journal of Finance, Vol. 2, pp. 41-55.
Grinblatt, M. and C.Y. Hwang (1989), Signaling and the Pricing of New Issues, Jour-nal of Finance, Vol. 44, No. 2, June 1989, pp. 393-420.
Huang, Q. and R.M. Levich (1998), Underpricing of New Equity Offerings by Privat-ized Firms: An International Test, Working Paper, Hofstra University and New York University, November 1998.
Jain, B.A. and O. Kini (1994), The Post-Issue Operating Performance of IPO Firms,
Journal of Finance, Vol. 49, No. 5, December 1994, pp. 1699-1726.
Jones, S.L. W.L. Megginson, R.C. Nash, J.M. Netter (1999), Share issue Privatiza-tions as financial means to political and economic ends, Journal of Financial Economics, Vol. 53, No. 2, pp. 217-253.
Keloharju, M. (1993), Winner’s curse, legal liability, and the long-run price perform-ance of initial public offerings in Finland, Journal of Financial Economics, Vol. 34, 1993, pp. 251-277.
Kothari, S.P. and J.B. Warner (1997), Measuring long-horizon security price performance, Journal of Financial Economics, Vol. 43, 1997, pp. 301-339. Kunz, R. M. and Aggarwal, R. (1994), Why initial public offerings are underpriced:
Evidence from Switzerland, Journal of Banking and Finance, Vol. 18, 1994. pp. 705-723.
Loughran, T. and J.R. Ritter (1995), The New Issues Puzzle, Journal of Finance, Vol. 50, No. 1, March 1995, pp. 23-51.
Loughran, T., J.R. Ritter and K. Rydqvist (1994), Initial public offerings: Interna-tional insights, Pacific-Basin Finance Journal, Vol. 2, 1994, pp. 165-199.
Lyon, J.D., B.M. Barber, and C.-L. Tsai (1999), Improved Methods for Tests of Long-Run Abnormal Stock Returns, Journal of Finance, Vol. 54, No. 1, Febru-ary 1999, pp. 165-201.
Megginson, W. L., R.C. Nash and M. Van Randenborgh (1994), The Financial and Operating Performance of Newly Privatized Firms: An International Empirical Analysis, Journal of Finance, Vol. XLIX, No. 2, June 1994, pp. 403-452.
Megginson, W.L., R.C. Nash, J.M. Netter and A.L. Schwartz (1998), The Long-Run Return to Investors in Share Issue Privatizations, Working Paper, University of Georgia and University of Oklahoma, Wake Forest University, University of Georgia and University of Miami, November 1998.
Menyah, K. and K.N. Paudyal (1996), Share Issue Privatisations: The UK Experience, in: Levis, M. (Editor), Empirical Issues in Raising Equity Capital, Elsevier, Am-sterdam 1996, pp. 17-48.
Paudyal, K., B. Saadouni and R.J. Briston (1998), Privatisation initial public offer-ings in Malaysia: Initial premuim and long-term performance, Pacific-Basin Fi-nance Journal, Vol. 6, pp. 427-451.
Perotti, E.C. (1995), Credible Privatization, American Economic Review, Vol. 85, No. 4, September 1995, pp. 847-859.
Perotti, E.C. and S.E. Guney (1993), The Structure of Privatization Plans, Financial
Management, Spring 1993, pp. 84-98.
Perotti E.C. and P. van Oijen (1999), Privatization, Political Risk and Stock Market Development in Emerging Economies, Working Paper, University of Amster-dam, CEPR and Tinbergen Institute, AmsterAmster-dam, March 1999.
Ritter, J.R. (1991), The Long-Run Performance of Initial Public Offerings, Journal of Finance, Vol. 46, No. 1, March 1991, pp. 3-27.
Rock, K. (1986), Why New Issues Are Underpriced, Journal of Financial Economics, Vol. 15, pp. 187-212.
Welch, I. (1989), Seasoned Offerings, Imitation Costs, and the Underpricing of Initial Public Offerings, Journal of Finance, Vol. 44, No. 2, June 1989, pp. 421-449.
Appendix
Table 1:
Descriptive statistics for the samples All Issues (All), Privatization IPOs (PIPOs), Private Sector IPOs
(IPOs) and Mass Privatization Program (MPP) as well as gross proceeds.
All PIPOs IPOs MPP
No of cases 149 51 98 15 e
Gross proceeds: Total (million PLN) 15,784.73 13,498.27 2,286.46 510.95
Mean (million PLN) 105.94 264.67 23.33
-Median (million PLN) 18.20 42.00 14.03
-Gross proceeds: Primary shares (million PLN) 2,950.35 890.87 2,059.48 0.00
% of total gross proceeds 18.7 6.6 90.0
Gross proceeds: Secondary shares (million PLN) 12,834.38 12,607.40 226.98 510.95
% of total gross proceeds 81.3 93.4 10.0 100.00
Fraction of capital sold in initial offer (%) 42.99 63.90 32.11
-Portion of issues: only secondary shares 29.5 66.7 10.2
-Portion of issues: only primary shares 57.0 3.9 84.7
-Portion of issues: secondary and primary shares 13.5 29.4 5.1
Portion of Firm Commitment Offerings 65.1 58.9 69.4
-Portion of Tender Offerings 34.9 41.1 30.6
-Mean Demand Multiple 2.30 2.75 2.07
-Average PE-Ratio at issue price 20.86 8.89 27.09 n.a.
Days from the start of the subscription period till the first trading day
87.32 61.26 100.88
-Mean fraction sold to employees (initial offer, %) 9.94 19.39 5.02
-e NIF certificate from 15.7.1996 till 11.6.1997 and 15 National Investment Funds shares (NIFs)
